The manuscript presents a technical setup for nitrate measurements in soil and open water in field conditions. Obviously, the subject is super important and critical in many scientific and environmental aspects. On one hand nitrogen pollution that is attributed to excess fertilization in agriculture is one of the main reasons for freshwater disqualification. On the hand the mechanisms related to the dynamic of nitrogen uptake and transport in the soil are also far from being well understood to achieve optimal fertilization. Therefore, any attempt to develop tools for filed measurement of nitrate in soil and water is important and should be published.

In general, the presented methodology involves a sophisticated setup which includes a set of 6 mini peristaltic pumps that synchronize several steps which are managed by an open-source controller.  Essentially the main steps include soil porewater sampling into a sampling tank where a reagent (that needs to be refrigerated in the field) is added to the sampled water until equilibrium is achieved (hours). Then the mixed sample is delivered into a cuvette for spectral analysis through synchronized automated process. In spite of its complexity and non-trivial field applicability the authors prove that it is doable and show reasonable results. Nevertheless, I have a major concern regarding the presented context and applicability of the proposed methodology.        

The main rational for developing the system is measuring soil nitrate concentration in agricultural setups. However, the measurement concentration range of the method presented here is inherently very low, limited to 10 ppm N-NO3 (which is the max allowed concentration in drinking water).  Nevertheless, soil nitrate concentration in fertilized agriculture is mostly far above that concentration, ranging between tens to hundreds ppm. Indeed, throughout their final chapter the author acknowledges this concentration limit. Nevertheless, I think that this limitation should be stated upfront in a clear manner, from the title and abstract to the objectives. Beyond this comment I want to state again that it is an elegant method that merit publication after revision that put in context the limited applicability.  

A very interesting paper looking at taking laboratory testing techniques out into the field and automating them.

There are several drawbacks to their current methods (which they do address in the paper, but perhaps could be expanded upon):

- low range - only 0-10 ppm NO3- - this is a very low nitrate level in an agricultural setting where we would expect levels a magnitude or two higher than this. The use of dilution to compensate for this is discussed.

- use of dry ice for cooling, perhaps a peltier cooling plate could be used (although this would have much higher power requirement)

- relies on soil-pore water being available - would not work in dry soil conditions. Measuring soil moisture/temperature as well would be beneficial.

- validation was only carried out over 5 weeks of field testing

References:

- include Bristow et al, 2022 in the discussion on LoRa sensing (section 3.2, 1a), as that work developed a remote LoRaWAN nitrate sensor

Overall a good, well written paper which presents a new automated field testing utilising existing laboratory techniques. It might be that this system is better suited to open water testing rather than in an agricultural situation, due to the requirement for water to sample and the low measurement range. It was good to see a github resource.